Marhsall is right, the RAAMmat should go on the door first, THEN the ensolite. However, I disagree that reversing the order nullifies the dampening effect. If you don't mind me getting a little nerdy, I'm pretty sure that in theory it doesn't matter what order you put these in with regards to each other. Any difference would be minor not major.
Why? Well I can't wrap my head around the mechanical problem as easily as I can do it with a simple R,L,C circuit. You can model any spring/mass system with and RLC circuit including a dampened system with a driving force such as the door. Take a look at the wikipedia page for the equivalent circuits:
http://en.wikipedia.org/wiki/Harmonic_oscillator#Equivalent_systems
In these equivalent circuits, the driving force is the vibration on the door metal (obviously caused by a number of things), the position x, and velocity dx/dt relate to the output, in this case the final vibrating surface which will, in turn be creating the sound we hear in the cabin, usually the inside of the cardboard/vinyl door panel.
To keep it simple, let's look at the series equivalent circuit. In that circuit the driving force (vibrating door metal) is Voltage and the output position is charge, and more usefully the velocity is current (I). The RAAMmat, if by itself would be one series connection of R (dampening), L (mass), and 1/C (springyness). So this would be a system where you just have RAAMmat by itself, and the sheet metal from the door is vibrating it. In order to add the ensolite, and the cardboard door panel, you need to add another R,L,andC in series for each one of these. Still following? I'll draw both circuits in crappy text drawings with their equivalents below:
Ground-----Voltage source----R----L----C-------R-----L----C--------R----L----C-------------------------Ground
Door metal RAAMmat Ensolite Carboard Door panel
Ground-----Voltage source----R----L----C-------R-----L----C--------R----L----C-------------------------Ground
Door metal Ensolite RAAMmat Carboard Door panel
Remember that the output, i.e. the sound vibration is actually the current running through the circuit, SO as you can see, the current would be the SAME in either one of these circuits. i.e. the noise coming out of the door will be the same, regardless of the order you place your dampeners.
Sure, there are non-idealities to the real-world senario, but I bet they aren't going to make more than a 10% difference in this case. After all, the door metal, the ensolite, and the RAAMmat are all in good physical contact. The cardboard door panel has a crappy mechanical contact, so I'm sure it would have sucked to put the RAAmat on the cardboard. I'm just pretty sure that the way I did it with the ensolite first, is actually working great as far as sound deadening.
Will it last as long ? I doubt it, the ensolite will peel off way before that RAAMmat goo ever looses it's tack. But I'm hoping it will last long enough to get me to the next time I need to tear that door panel apart to get to something.